Research On Fault Detection Method Of Mechanical System Based On Broadband Spectrum Processing

On-line monitoring of the running state and faults of mechanical systems and timely detection of recessive and early faults have always been a research hotspot at home and abroad.The traditional mechanical system fault detection technology with certain latent fault detection ability is mainly based on narrow-band vibration analysis technology,which mainly detects faults and determines the type by analyzing the characteristic frequency corresponding to faults and its multiple frequency.However,these technologies have insufficient bandwidth to pick up and process fault information,and can only find and judge the typical mechanical equipment faults that have occurred and are relatively serious,while the ability to predict some hidden faults of equipment is relatively inadequate.In the case of strong background noise,the composition of vibration signals is often more complex,and it is easy to cause the aliasing of fault signals in the demodulation process,which makes it difficult to distinguish the characteristic frequency and its multiple frequency of fault signals.Therefore,the narrow-band vibration analysis technology has been unable to meet the requirements of full automation and intelligent online monitoring of mechanical system running state and fault.Based on this,the next generation fault detection technology using acoustic sensors based on broadband spectrum processing has been studied abroad,but the public literature is still few..Sound is generated by vibration,and both of them are mechanical waves.However,when the mechanical system is in the early stage and the hidden fault state,the broadband spectrum processing can be realized by using the good broadband information picking ability of acoustic sensors.The ability of early fault detection and the accuracy of type determination can be significantly enhanced by making full use of the broadband information of fault signals.This paper presents a fault detection method of mechanical system based on broadband spectrum processing.This method is mainly based on acoustic signal picking and processing,through to the mechanical equipment under the normal operation of the acoustic signal acquisition,tracking,and background noise suppression,establish a sound system normal grain base,and then the voiceprint matching based on broadband spectrum processing and pattern recognition technology,realize the homologous between signals and the homologous detection and classification,and then realize the mechanical system running state and implicit effective monitoring and fault type.The main work of this paper is as follows:(1)study a mechanical system fault detection method based on wideband spectrum processing,the method adopts the broadband voiceprint background noise suppression and broadband voiceprint matching technology,acoustic signal processing on mechanical equipment running status,through the same voiceprint,signal detection and classification,implementation of mechanical system running state monitoring and fault type..(2)The proposed fault detection method of mechanical system based on broadband spectrum processing is studied by computer numerical simulation,and compared with the traditional fault detection method.(3)Based on Lab VIEW development platform,the fault monitoring prototype system integration was realized,and the prototype system was preliminarily tested.According to the test results,the prototype system can monitor the state change of voiceprint signal in real time.